Electric convulsion therapy has been generally accepted as a valuable method of treatment in depressive states. Assessment of the results of this form of treatment has generally been based on a small series of cases, and percentage rates of recovery and relapse have usually been computed on one hospitalization. No report has so far been published of a complete follow-up of patients throughout their frequent hospitalizations and remissions. Nevertheless, it is only on such a long-term basis that treatment of depressive states can be accurately evaluated. It was with this ideal in mind that the present study was undertaken. The clinical cases on which it is based are those of 923 patients treated in the Royal Edinburgh Hospital for Mental and Nervous Disorders between the years 1930-48.